A Variable Neighborhood Search-Based GIS for Real-Time Health Service Distribution to Disabled and Chronic Patients in Disaster and Emergency Scenarios

Abstract

This study presents a real-time Geographic Information System (GIS) enhanced with a Variable Neighborhood Search (VNS) algorithm to optimize the distribution of health services for disabled and chronically ill patients during disasters. The system solves a multi-vehicle routing problem with time windows (VRPTW), where patient priority is dynamically assigned based on medical urgency and vulnerability. The framework integrates Google Maps APIs for geospatial services, a full-stack application built with Node.js and React, and MongoDB for data management. The VNS-based optimization module minimizes total travel distance and vehicle count while prioritizing critical cases. Experiments using an augmented Solomon C101 dataset show that the proposed method reduces total distance by 18.7% and vehicle requirements by 21.4%, while serving high-priority patients 32% faster compared to Genetic Algorithm and First-Come-First-Served benchmarks. The research contributes a functional web-GIS platform for real-time emergency coordination and demonstrates the effectiveness of metaheuristic VNS in dynamic healthcare logistics. The system offers a scalable decision-support tool to improve response equity and efficiency in crisis scenarios. Future work will integrate IoT patient monitoring and multi-objective optimization.